[Update 16.09 GMT: Additional details added on Facebook's roadmap]
In the next decade, Facebook plans to go beyond social networking in order to boost global internet connectivity, develop artificial intelligence technology, and make virtual reality accessible to everyone.
Speaking at the technology conference Web Summit 2016 in Lisbon, Portugal, Mike Schroepfer, Facebook's chief technology officer (CTO), revealed Facebook's three long-term goals, set out to take place over the next 10 years.
The executive said that Facebook's 10-year-old vision centers on "the most important problems we are trying to solve."
The first problem is how to connect everyone around the world to the internet. It is estimated that 4.1 billion people have no access to the web, and Facebook wants to make sure "their voices are brought to the internet."
In order to do so, Facebook wants to explore how to expand internet connectivity through both rural and urban areas. The tech giant is not only experimenting with satellites -- despite the failure of the AMOS 6 satellite due to bring the internet to underserved parts of Africa and being blown up instead in the recent SpaceX disaster -- but wireless technology which does not require miles of cables to be laid.
One particular project which Facebook is developing is the creation of a plane able to fly at high altitudes for long periods of time. These solar-powered planes are made of carbon fiber alloys and have the wingspan of a Boeing 737. Each one weighs less than 1000 pounds.
Facebook says these unmanned aerial vehicles (UAVs) are designed to fly "for months at a time" and teams of these planes will eventually fly to provide suburban area internet connectivity.
Rather than laying cables, these UAVs will provide a "backbone in the sky" for web connectivity, according to Schroepfer.
These UAVs are due to be tested soon in San Jose, California. In the meantime, the technology behind the endurance planes is being made available as part of the open-source Open Compute project and Telecom Infra project. By releasing this research to the open-source community, Facebook says that this can "dramatically reduce the cost of providing internet access across the world."
Facebook's second long-term goal is to build "truly intelligent machines capable of managing the mass of information we post online," according to Schroepfer.
While artificial intelligence (AI) is already used by Facebook in image recognition and facial detection processes, the company says that AI usage does not have to purely focus on knowledge, but also "for creation."
The company hopes that it can eventually create machines which have the same creative spark as humanity. The same neural nets used by us to be creative, perhaps, one day will be used in AI, in order to help us create everything from art to applications in the medical industry.
"Our work in AI is helping us move all these projects forward," Facebook says. "We're conducting industry-leading research to help drive advancements in AI disciplines like computer vision, language understanding, and machine learning. We then use this research to build infrastructure that anyone at Facebook can use to build new products and services."
FBLearner Flow, for example, is considered the "backbone" of AI-based product development at Facebook. The platform is used for AI experiements by staff who do not need in-depth knowledge of the field to run tests on new products.
In addition, the tech giant has developed AutoML, an engineering platform which improves new AI systems based on data from existing AI products. The company says that over 300,000 machine learning models are being tested and optimized every month thanks to the platform.
One of Facebook's newest AI additions is called Lumos, a "self-serve" platform which gives Facebook teams the chance to harness artificial intelligence algorithms for products and services without prior knowledge. The company says that Lumos has "improve[d] our ability to spot content that violates our community standards."
This is not the end of the social network giant's goals, however. Facebook also believes that there is an "exciting future" in shifting these neural networks from servers to devices in our pocket, making AI available to all for everything from creative media editing to voice assistants.
Facebook has made headway in the last three months through the development of Caffe2Go, a deep learning platform which is able to able to use AI and ML to doctor images and video in real time. The app captures, analyzes, and processes pixels through AI, made possible by condensing the size of the artificial intelligence model used to process content by 100x.
In the next 10 years, Facebook also plans to promote the research and development of virtual reality (VR) technology. The executive says that Facebook ultimately wants to make VR technology "cheaper, easy to use, and highly distributed," and 2016 is the year to do it as component technology "has caught up" with the company's software -- and plans.
"Like AI, we needed to wait for all the component technology to catch up with us," the executive said.
According to Schroepfer, in order to reach these lofty goals for a technology which is still in its infancy when it comes to the consumer market, the tech giant is developing a VR device which does not need any external equipment or a PC to function.
Facebook's headset, simply dubbed Standalone for now, carries cameras on board which track the position of the user so movements coincide with the virtual world. In addition, the company is working on Touch, a set of virtual reality controllers which focus on hand movements.
The devices are currently in prototype stage and will not be on the open market anytime soon.
"The future is the ability to connect with the people you care about no matter whether you are a few or hundreds of miles away [...] and experience anything with anyone at any time."
Disclaimer: Web Summit 2016 sponsored the trip to Lisbon, Portugal.